How to Organise Nonprofit Evidence for Grant Proposals and Impact Reports

When a colleague needs a result for a grant proposal, can they find the supporting record and explain what the number means? If every request has to go back to the person who collected the information, a folder of reports is only doing part of the job. Your team also needs a way to identify the right result, check its meaning and know where it can be used.

An evidence library is an index of your programme data and findings, linked to the records that support them. For a non-profit team, it can begin with a shared spreadsheet and the files it references. The spreadsheet explains the information; the files preserve the supporting records. The arrangement matters when programme staff, fundraising staff and communications staff need different outputs from the same work.

A grant proposal may use a previous result to explain the case for further work. An impact report needs to account for what happened during a defined period, including what was learned. Organising evidence well allows both writers to begin from an agreed account and develop it for their readers.

This guide shows how to build that arrangement around one frequently used result. We will choose the claim, record its definition and source, and separate restricted records from information cleared for communication. A worked health-service example then shows how the same evidence becomes two different paragraphs: one for a proposal and one for an impact report. Finally, we will follow a correction through the records and documents affected by it.

You can begin with the files you already have. The first decision is which statements your team needs to be able to substantiate.

Choose the programme results your evidence library needs to document

Begin with one programme and the documents your team already produces about it. Identify the statements that recur: how many people used a service, what changed for participants, what the organisation delivered, what it spent and what it learned. These statements give the first entries a purpose.

A large collection can wait. Working through a small set of frequently used claims will reveal which records are usable and which need clarification. Include an inconvenient finding if it affects how the programme should be understood. A library assembled only from favourable results will be a poor starting point for learning or reporting.

NPC's principles of good impact reporting describe an internal as well as an external purpose: reporting helps stakeholders understand the work and helps staff and trustees focus on results. For the evidence library, that means retaining the information needed to explain a decision, including findings that prompted the team to change its approach.

Record what each claim measures before copying the number. A count of clinic visits and a count of people served can both be correct while differing substantially. One person may attend several times. The writer needs the definition to choose the right statement, and the programme team needs it to make a meaningful comparison over time.

If the difficulty is deciding what evidence to collect in the first place, Ije's Proof Evidence Capture Planner can help organise a starter capture plan. For evidence already available, the next step is to make each selected result traceable.

How to organise programme data in an evidence library

Create an entry for a claim that someone is likely to reuse. Give it an identifier that remains stable when the wording changes. Then record the source, the period and population covered, the calculation or method, and the person who can explain it.

Example: an evidence-library entry for a health information service

Here is an illustrative entry for a community health information service. The service, records and figures are constructed for this example. The entry demonstrates how a writer could trace a public statement; it is not an account of an Ije client.

The definition explains why the word “requests” matters. Replacing it with “people” would change the claim even if the number stayed the same. Deduplication here removes duplicate log entries, not repeat users with different enquiries.

Use a direct file link in the real entry and identify the relevant worksheet, page or table. A link to a large folder still leaves the writer to find the evidence. If a summary calculation supports the statement, link that calculation as well as the underlying record so a reviewer can retrace the total.

The fields above are a practical starting arrangement proposed here. They are not a universal reporting standard. A qualitative interview, an evaluation finding and a spending figure need different details. Keep the information necessary to understand and check each claim, rather than requiring every source to fit the same numerical template.

Separate confidential programme records from publishable evidence

Programme records may contain names, contact details, health information or accounts that were never intended for public communication. A communications writer often needs a verified aggregate or an agreed excerpt rather than access to every underlying record.

Use the existing storage system's access controls to distinguish restricted records, reviewed summaries and public outputs. The evidence index should identify the source custodian and point the writer to the version they are authorised to use. This also gives them a route for questions when they cannot inspect the raw record directly.

For the illustrative health-service entry, the programme team could maintain the restricted register while sharing a summary that contains the quarterly count and its definition. The public article or report could use that summary without reproducing individual enquiries. A participant story would require a separate decision about permission, context and attribution; clearance of an aggregate total would not cover the story.

Keep public wording separate from source notes. A short sentence suitable for a website may omit methodological detail that remains important to an evaluator. Retain that detail in the entry and decide what each reader needs. The non-profit brand storytelling guide discusses the additional considerations when a person's account is part of the material.

With a checked result and permission to use it, the writer can concentrate on what the reader needs to understand. This is where the proposal and the impact report take different approaches.

How to use programme evidence in grant proposals and impact reports

The evidence stays consistent while the explanation changes with the document's purpose. A proposal uses it to support a reasoned plan. A report uses it to account for completed work and learning.

Continue the illustrative health-service example. Suppose a review of the 84 requests found that 31 concerned how to book appointments. Assume the team recorded this classification consistently and retained the review notes. The team now proposes a plain-language appointment guide, checked with the relevant service providers.

Grant proposal example: explaining the case for new work

A proposal paragraph could say:

During the first quarter of 2026, our information service answered 84 requests. Of these, 31 concerned how to book appointments. We propose developing a plain-language appointment guide with local service providers and testing whether readers can use it to identify the correct booking steps. This responds to a recurring question recorded by the service. The project budget will cover development, provider review and user testing.

Impact report example: explaining activity and learning

A report paragraph using the same evidence could say:

The information service answered 84 requests between January and March 2026. Appointment booking accounted for 31 requests. Following our review of these enquiries, we decided to develop an appointment guide with local service providers. We have not yet assessed whether the guide helps people book appointments; development and testing are the next stage.

The proposal explains why a new activity is justified and what it will involve. The report explains the period's activity and the decision made from it. Neither turns the volume of enquiries into a claim about improved health or the number of people served.

Both paragraphs can reference HEALTH-2026-01 and a second entry for the enquiry classification in the team's working files. A later test of the guide would create new evidence. It should not silently change the meaning of the first-quarter result.

Correct programme data in reports, proposals and website content

When a source changes, identify the outputs affected by it. Alongside the evidence entry, keep a short list of material uses: a named proposal, a report page or a website URL. This lets the team find a statement that needs correction without searching every file it has ever produced.

Suppose, in the illustrative case, a subsequent review finds one further duplicate enquiry. The verified total becomes 83. The coordinator records the correction, its date and the reason; preserves the earlier version; and checks whether the appointment-booking subtotal also changes. The communications lead then updates the relevant draft and checks any published use. Where an already distributed document materially misstates the work, decide how to notify its recipients and identify the correction.

Bond's Evidence Principles are designed to help NGOs assess evidence quality in research, evaluations and case studies. They support reviewing the underlying evidence before sign-off. The correction process described here applies that concern to reuse: someone must be able to see why the statement changed, rather than finding a different total with no explanation.

Schedule review around events that can change the record, such as the end of a reporting period, a completed evaluation or a correction from programme staff. Recheck an entry before a substantial new use. A corrected total may need a short verification; a disputed interpretation may require the evaluator or programme lead to examine the underlying work.

Build and test your first evidence-library entry

Choose a claim from a recent report and build its entry. Ask a colleague who did not prepare the report to locate the source and explain the number in their own words. If they cannot tell what it measures, improve the definition. If the source is restricted, provide an authorised summary and a named person who can answer questions. If the wording suggests more than the record establishes, revise it before reuse.

The entry is usable when someone can produce an accurate paragraph from it and another person can check that paragraph against the evidence. Add the next frequently used result once that process works. If you need to assess the wider collection, Ije's Proof Bank Readiness Diagnostic can help identify missing preparation before you expand it. The diagnostic does not verify the underlying evidence; that remains a source-review task.

A shared evidence library gives programme and communications staff an agreed starting point for explaining the organisation's work. It reduces the need to reconstruct a result from memory and makes differences between documents easier to examine. The organisation can then adapt its explanation to a funder, participant or public reader while preserving what it knows about its own work.